of misconduct in science in governmental and institutional policies and procedures can lead to uncertainty about whether to include as misconduct in science those cases that involve charges of incompetence, science conducted with “reckless disregard” for the truth, and other misconduct, such as sexual harassment, that may occur in the research environment. 17
Nature of evidentiary findings. University and government officials sometimes differ on the nature of evidence that is necessary to substantiate an allegation or suspicion of misconduct in science. For example, some institutions have concluded that carelessness and poor judgment do not constitute misconduct in science. Government officials have sometimes disagreed with such findings, particularly when, in the government's view, there was evidence to show that deception was intentional. In other cases, government officials have criticized or rejected institutional reports of inquiries or investigations as “defective” when these reports lacked sufficient information to enable others to assess the fairness or completeness of the institutional process.
There are also differences in evidentiary standards used by research institutions and government agencies. Institutional officials may prefer to use a higher standard of “clear and convincing” evidence or evidence that is “beyond a reasonable doubt,” while government officials may rely on a less restrictive “preponderance-of-the-evidence” standard to substantiate a finding of misconduct in science.
Due process requirements. There is confusion about the formal procedures that are required in the resolution of allegations of misconduct in science. Since government officials often rely on institutional investigative reports in recommending possible sanctions, there can be different expectations and standards of procedural clarity, fairness, and objectivity. The OSI's approach, which has been criticized (Hamilton, 1991a), maintains what it calls the “scientific dialogue” model of investigation, in contrast to what OSI staff term the “legal-adversarial” approach adopted by NSF, in the belief that the former method can both reveal the scientific facts of the case and also secure the due process rights of the respondent without inviting the difficulties of adversarial proceedings.18
Some of the policies and procedures used by OSI in its investigations and oversight have been challenged in the courts and criticized in the press. Several problems have been identified: the inability of the subjects and key witnesses of the investigations to review all evidence until the compilation of the investigative report, premature disclosure of draft reports in the press, and the absence of disclosure